CN113176460A - Power quality disturbance signal detection method based on improved empirical wavelet transform - Google Patents
Power quality disturbance signal detection method based on improved empirical wavelet transform Download PDFInfo
- Publication number
- CN113176460A CN113176460A CN202110464981.5A CN202110464981A CN113176460A CN 113176460 A CN113176460 A CN 113176460A CN 202110464981 A CN202110464981 A CN 202110464981A CN 113176460 A CN113176460 A CN 113176460A
- Authority
- CN
- China
- Prior art keywords
- frequency
- disturbance
- signal
- power quality
- disturbance signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 claims abstract description 39
- 238000001228 spectrum Methods 0.000 claims abstract description 25
- 238000005259 measurement Methods 0.000 claims abstract description 10
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000000354 decomposition reaction Methods 0.000 abstract description 19
- 230000000694 effects Effects 0.000 abstract description 7
- 239000000284 extract Substances 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 22
- 230000007547 defect Effects 0.000 description 5
- 230000009466 transformation Effects 0.000 description 5
- 238000010276 construction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 239000003595 mist Substances 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 210000002816 gill Anatomy 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Noise Elimination (AREA)
Abstract
The invention discloses a power quality disturbance signal detection method based on improved empirical wavelet transform, which comprises the following steps: carrying out scale space representation on the frequency spectrum of the power quality disturbing signal, and determining the characteristic frequency point of the disturbing signal according to the dynamic measurement of the scale space representation result; extending the boundary on the basis of the original frequency band boundary, and constructing an experience wavelet on the basis of the extended frequency band boundary according to the acquired characteristic frequency points of the disturbance signal; decomposing the power quality disturbance signal into a plurality of amplitude modulation-frequency modulation components by using improved empirical wavelet transform; and acquiring a detection result of the disturbance signal according to the amplitude modulation-frequency modulation component. The method effectively inhibits the phenomenon of spectrum 'over-segmentation' of the traditional EWT, simultaneously has no modal aliasing problem, fully considers the decomposition result characteristic of the IEWT, and adopts an NHT method to accurately extract the disturbance parameters from each mode, thereby having good disturbance time-frequency detection effect.
Description
Technical Field
The invention belongs to the field of power system automation, relates to a power quality disturbing signal detection and analysis technology, and particularly relates to a power quality disturbing signal detection method based on improved empirical wavelet transform.
Background
With the continuous promotion of smart grid construction, the number of power electronic devices connected into a power system is continuously increased, the source load fluctuation characteristics are increasingly remarkable, and a series of serious power quality pollution problems are caused. Meanwhile, the wide use of precision instruments and sensitive equipment also puts higher requirements on the power quality of the power grid. In order to effectively manage and evaluate the power quality problem, various power quality disturbances are firstly dealt with for rapid and accurate analysis.
In recent years, researchers at home and abroad have conducted intensive research on a power quality disturbance detection method. Short Time Fourier Transform (STFT) can detect a disturbance signal in combination with Time-frequency information, but since a window function is fixed, detection accuracy needs to be improved; wavelet Transform (WT) overcomes the defect of single STFT time-frequency resolution, but Wavelet basis and decomposition layer number are difficult to determine; the S transformation is used as the extension of STFT and WT, has good time-frequency analysis performance and noise robustness, but the analysis result is still limited by the Heisenberg inaccurate measurement principle, and the operation is complicated; non-stationary signal Decomposition algorithms represented by Hilbert Transform-Huang (HHT) and Local Mean Decomposition (LMD) are also applied to power quality disturbance analysis and show good adaptivity, but the methods have the defects of modal aliasing, noise sensitivity and the like, and the application range of the methods is severely limited. In addition, Variational Mode Decomposition (VMD) is also applied to power quality disturbance detection, the method well overcomes the modal aliasing phenomenon, but algorithm parameters need to be preset, adaptive detection cannot be realized, and the detection effect is greatly influenced by the parameter size; singular value gradient information is also used for detecting transient power quality disturbance parameters, and the method has good noise suppression capability and zero crossing point detection effect, can accurately position disturbance time, but cannot detect the disturbance parameters such as amplitude, frequency and the like.
Empirical Wavelet Transform (EWT) is a new signal analysis method proposed by Gilles, a french learner, in 2013, and the method integrates the adaptivity of Empirical Mode Decomposition (EMD) and the theoretical framework of Wavelet analysis, can obtain a good analysis effect with a small amount of operation cost, overcomes the defect of fixed division of WT frequency bands while suppressing the problem of modal aliasing of EMD, and has been effectively applied to engineering signal analysis. However, the existing method cannot effectively solve the problem of spectrum "over-slicing" existing in the conventional EWT, which causes the disadvantage of weak adaptivity when the EWT is applied to the analysis of the power quality disturbance signal.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the defects of the existing EWT technology in the field of electric energy quality signal analysis and the defects of the existing electric energy quality disturbing signal detection method in the aspect of detection precision, the electric energy quality disturbing signal detection method based on Improved Empirical Wavelet Transform (IEWT) is provided, has good modal decomposition capability and anti-noise performance, is suitable for the detection and analysis of complex electric energy quality disturbing signals, and is used for solving the problem that when the Empirical Wavelet Transform (EWT) is used for electric energy quality signal analysis, the frequency band division result is easily interfered by frequency spectrum leakage and noise pollution.
The technical scheme is as follows: in order to achieve the above object, the present invention provides a method for detecting a power quality disturbance signal based on improved empirical wavelet transform, comprising the following steps:
s1: carrying out scale space representation on the frequency spectrum of the power quality disturbing signal, and determining the characteristic frequency point of the disturbing signal according to the dynamic measurement of the scale space representation result;
s2: extending the boundary on the basis of the original frequency band boundary, and constructing an experience wavelet on the basis of the extended frequency band boundary according to the acquired characteristic frequency points of the disturbance signal;
s3: decomposing the power quality disturbance signal into a plurality of amplitude modulation-frequency modulation components by using improved empirical wavelet transform;
s4: and acquiring a detection result of the disturbance signal according to the amplitude modulation-frequency modulation component.
Further, the step S1 specifically includes the following steps:
a1: setting an electric energy quality disturbing signal containing N characteristic frequency points as F (t), and performing fast Fourier transform on the signal to obtain a result F (omega);
a2: a discrete form of Scale-Space Representation (SSR) that defines F (ω);
a3: and carrying out trend removing evaluation on the maximum value point in the scale space representation by adopting dynamic measurement, and determining the characteristic frequency point of the disturbance signal.
Further, the Gaussian kernel function defined in the step A2 is
In the formula: n is a scale factor; the discrete form of the scale-space representation of F (ω) is then:
in the formula: n is a scale factor, M is an approximation coefficient, when
The discrete approximation error of the gaussian kernel is negligible. The SSR can keep the frequency spectrum characteristic information and simultaneously carry out smoothing treatment on the local interference peak value, so that the distinguishing difficulty of the characteristic frequency point can be obviously reduced.
Further, the step a3 is specifically: let xuIs a maximum point in L (ω, n) if there is a ratio x in L (ω, n)uHigher points, then xuDynamic measure D ofyn(xu) Equal to its dynamic measure to the smallest path among all the equal altitude point paths, i.e.:
Dyn(xu)＝{inf{Dyn[l(xu,xv)]}；halt(xu)＝halt(xv)}
in the formula: l (x)u，xv) Denotes xuTo the isocenter xvA path of (a); i.e. inf{. denotes infimum bound; h isalt(. cndot.) represents height; at this time, if Dyn(xu) Greater than a discrimination threshold TdCan determine xuIs a characteristic frequency point.
Further, the specific process of step S2 is as follows:
b1: capture all N characteristic frequency points { omeganAfter the Nn is equal to 1, selecting a frequency corresponding to a midpoint between two adjacent characteristic frequency points as a frequency band division boundary; at the same time, the first N boundaries ω are preservednAnd (5) extending the band boundary according to the following formula on the basis that N-1N is 0:
in the formula: lambda is an adjustable boundary extension coefficient; at this time, the signal spectrum is divided into N +1 continuous intervals
B2: defining an empirical wavelet tightening framework
And constructing an empirical scale function in the frequency domain
And empirical wavelet function
Respectively performing inner products on f (t) and empirical scale function and empirical wavelet function by using a method similar to classical wavelet transform to obtain approximate coefficients
And detail coefficient
Further, in the step B1, according to the number of characteristic frequency points in the signal spectrum, the power quality disturbance is divided into a single-frequency disturbance and a multi-frequency disturbance; for pure single-frequency power quality disturbance, referring to a frequency band subdivision rule defined by a Mallat algorithm, setting lambda (pi/2-omega)N)/(ΩN-ωN-1) I.e. define the interval ΛN+1＝[π/2，π]Obtaining high-frequency distortion information contained in a disturbance signal frequency spectrum; when the disturbance signal is mixed with noise interference, considering that the disturbance component frequency spectrum is symmetrical about a peak point in a frequency domain, making λ ═ 1 can make the highest frequency component completely fall into an upper cut-off frequency and a lower cut-off frequency which are respectively ω while keeping the frequency domain characteristics of the disturbance signalN-1And ωNThereby being free from high frequency noise.
Further, the power quality disturbance signal in the step S3 is a voltage signal or a current signal.
Further, the step S4 is specifically: and (2) performing standard Hilbert transform (NHT) on each disturbance component to obtain the instantaneous amplitude and the instantaneous frequency of the disturbance component, so as to obtain a time-frequency spectrogram of the disturbance signal, the disturbance amplitude, the disturbance frequency and disturbance start-stop time parameters.
Further, the method for calculating the instantaneous amplitude and the instantaneous frequency of the disturbance component in step S4 includes:
c1: fitting the maximum value point of | x (t) | by adopting a cubic spline function to obtain an empirical envelope function a (t), and standardizing the single-component signal x (t) according to the following formula:
x1(t)＝x(t)/a(t)
c2: obtaining a normalized signal | x1(t) | empirical envelope function a1(t) in theory at this time a1(t) should be less than or equal to 1, otherwise the above steps are iterated n times until an(t) is less than or equal to 1; to this end, the single component x (t) has been decomposed into the form of a product of amplitude modulation and frequency modulation, the resulting signal x being normalizedn(t) is the frequency modulated signal, and the amplitude modulated portion A (t) is defined as:
A(t)＝x(t)/xn(t)＝a(t)a1(t)…an(t)
c3: at this time, xn(t) already for an approximately pure frequency-modulated signal, its instantaneous frequency function f (t) is obtained directly by the hilbert transform:
compared with the conventional Hilbert transform result, the decomposition result of the IEWT in step S4 of the invention has more stable amplitude at the endpoint and no flying wing phenomenon, and is more suitable for extracting disturbance information in the IEWT decomposition result.
Has the advantages that: compared with the prior art, the invention has the following advantages:
(1) the invention effectively inhibits the phenomenon of spectrum 'over-segmentation' of the traditional EWT, and simultaneously has no modal aliasing problem.
(2) The invention fully considers the decomposition result characteristic of the IEWT and adopts the NHT method to accurately extract the disturbance parameters from each mode, thereby having good disturbance time-frequency detection effect.
(3) The invention does not need a pre-filtering unit, can accurately analyze pure or noise-contaminated disturbance signals, can obtain good detection effect on zero-crossing point disturbance, and has strong universality.
(4) Compared with other time-frequency analysis methods, the method has the advantages of low computation amount, good real-time performance, easy integration to a digital monitoring device and suitability for rapid and accurate detection of the power quality disturbance signal.
Drawings
FIG. 1 is a flow chart of an improved empirical wavelet transform;
FIG. 2 is a graph of voltage ramp waveforms and their spectral division results;
fig. 3 is a graph of the voltage sag disturbance analysis result (SNR ∞ dB);
fig. 4 is a graph of the voltage sag disturbance analysis results (SNR ═ 20 dB);
FIG. 5 is a graph of measured harmonic perturbations and their spectral partitioning results;
FIG. 6 is a graph of the results of an analysis of measured harmonic perturbations.
Detailed Description
The present invention is further illustrated by the following figures and specific examples, which are to be understood as illustrative only and not as limiting the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalent modifications thereof which may occur to those skilled in the art upon reading the present specification.
The invention provides a power quality disturbance signal detection method based on improved empirical wavelet transform, which comprises the following steps:
s1: and performing scale space representation on the frequency spectrum of the power quality disturbing signal, and determining the characteristic frequency point of the disturbing signal according to the dynamic measurement of the SSR result.
The method specifically comprises the following steps:
a1: setting an electric energy quality disturbing signal containing N characteristic frequency points as F (t), and performing fast Fourier transform on the signal to obtain a result F (omega);
a2: defining a Gaussian kernel function of
In the formula: n is a scale factor; the discrete form of the scale-space representation of F (ω) is then:
in the formula: n is a scale factor, M is an approximation coefficient, when
The discrete approximation error of the gaussian kernel is negligible. The SSR can keep the frequency spectrum characteristic information and simultaneously carry out smoothing treatment on the local interference peak value, so that the distinguishing difficulty of the characteristic frequency point can be obviously reduced.
A3: the SSR definition shows that the interference bins will disappear gradually as n increases, but the characteristic bins will be submerged. In order to obtain higher frequency domain resolution of the characteristic frequency points and simultaneously consider that interference extreme points with larger amplitude possibly exist in the SSR under a smaller scale, dynamic measurement is adopted to perform trend removing evaluation on the maximum value points in the SSR in the step.
Let xuIs a maximum point in L (ω, n) if x is present in L (ω, n)uHigher points, then xuDynamic measure D ofyn(xu) Equal to its dynamic measure to the smallest of all the paths of the isocenter, i.e.
Dyn(xu)＝{inf{Dyn[l(xu,xv)]}；halt(xu)＝halt(xv)}
In the formula: l (x)u，xv) Denotes xuTo the isocenter xvA path of (a); i.e. inf{. denotes infimum bound; h isalt(. cndot.) represents height. At this time, if Dyn(xu) Greater than a discrimination threshold TdCan determine xuIs a characteristic frequency point. In this embodiment, n is 9, Td＝5。
S2: and extending the boundary on the basis of the original frequency band boundary, and constructing an empirical wavelet on the basis of the extended frequency band boundary according to the acquired characteristic frequency points of the disturbance signal. The method specifically comprises the following steps:
b1: capture all N characteristic frequency points { omeganAfter the Nn is equal to 1, selecting a frequency corresponding to a midpoint between two adjacent characteristic frequency points as a frequency band division boundary; at the same time, the first N boundaries ω are preservednAnd (5) extending the band boundary according to the following formula on the basis that N-1N is 0:
in the formula: lambda is an adjustable boundary extension coefficient; at this time, the signal spectrum will be divided into N +1 consecutive intervals { Λ +n}N+1n＝1；
According to characteristic frequency in signal frequency spectrumThe number of points, the power quality disturbances can be divided into single frequency disturbances and multiple frequency disturbances. For pure single-frequency power quality disturbance, referring to a frequency band subdivision rule defined by a Mallat algorithm, setting lambda (pi/2-omega)N)/(ΩN-ωN-1) I.e. define the interval ΛN+1＝[π/2，π]So as to obtain the high-frequency distortion information contained in the disturbance signal spectrum. When the disturbance signal is mixed with noise interference, considering that the disturbance component frequency spectrum is symmetrical about a peak point in a frequency domain, making λ ═ 1 can make the highest frequency component completely fall into an upper cut-off frequency and a lower cut-off frequency which are respectively ω while keeping the frequency domain characteristics of the disturbance signalN-1And ωNThereby being free from high frequency noise.
B2: defining an empirical wavelet tight framework with reference to Littlewood-Paley and Meyer wavelet construction methods
And constructing an empirical scale function in the frequency domain
And empirical wavelet function
Respectively performing inner products on f (t) and empirical scale function and empirical wavelet function by using a method similar to classical wavelet transform to obtain approximate coefficients
And detail coefficient
In addition, when the window width factor γ satisfies the following formula, a corresponding tight-bracket window function can be obtained, and the value of the tight-bracket window function has no significant relation to the disturbance detection effectmin(1-1/L)/2, L is the length of the perturbation signal.
The improved empirical wavelet transform implementation flow is shown in fig. 1.
S3: the improved empirical wavelet transform is used for decomposing the power quality disturbance signal into a plurality of amplitude modulation-frequency modulation components, wherein the disturbance signal can be a current signal or a voltage signal.
S4: and (3) performing standard Hilbert transform on each disturbance component to obtain the instantaneous amplitude and instantaneous frequency of the disturbance component, so as to obtain a time-frequency spectrogram of the disturbance signal, and parameters of the disturbance amplitude, the disturbance frequency and the disturbance starting and stopping time.
The method for solving the instantaneous amplitude and the instantaneous frequency of the disturbance component comprises the following steps:
c1: fitting the maximum value point of | x (t) | by adopting a cubic spline function to obtain an empirical envelope function a (t), and standardizing the single-component signal x (t) according to the following formula:
x1(t)＝x(t)/a(t)
c2: obtaining a normalized signal | x1(t) | empirical envelope function a1(t) in theory at this time a1(t) should be less than or equal to 1, otherwise the above steps are iterated n times until an(t) is less than or equal to 1; to this end, the single component x (t) has been decomposed into the form of a product of amplitude modulation and frequency modulation, the resulting signal x being normalizedn(t) is the frequency modulated signal, and the amplitude modulated portion A (t) is defined as:
A(t)＝x(t)/xn(t)＝a(t)a1(t)…an(t)
c3: at this time, xn(t) already for an approximately pure frequency-modulated signal, its instantaneous frequency function f (t) is obtained directly by the hilbert transform:
compared with the traditional Hilbert transform result, the IEWT decomposition result has the advantages that the amplitude of the standard Hilbert transform result at the end point is more stable, the flying wing phenomenon does not exist, and the IEWT decomposition result is more suitable for extracting disturbance information in the IEWT decomposition result. Therefore, the single component x (t) obtained by IEWT decomposition is demodulated by adopting NHT (non-uniform time transform) to obtain a better power quality disturbance time-frequency analysis effect; meanwhile, a disturbance signal time-frequency spectrogram is obtained and the detection of the amplitude, the frequency and the start-stop moment of the disturbance signal is realized by combining the transformation result of the IEWT and the instantaneous amplitude and the frequency function of each disturbance component.
Based on the above scheme, this embodiment applies the method of the present invention as an example, specifically as follows:
example 1:
for the voltage temporary rise disturbance f with the sampling frequency of 6.4kHz and the signal duration of 0.24s1(t) start and stop moments of the pause are 0.0652s and 0.1645s, respectively, and the IEWT is used for f1(t) detection is carried out, the disturbance waveform and the frequency band division result thereof are shown in FIG. 2, and the disturbance analysis result is shown in FIG. 3. As can be seen from this, for a clean voltage-transient disturbance signal, the number N of characteristic frequency points is 1, so the original signal is decomposed into an approximate component c1And a detail component c2. As can be seen from the observation of FIG. 3, in the approximate component c1While preserving the transient-rise disturbance amplitude characteristic, the IEWT also inherits the good singular point detection performance of the classical wavelet transformation according to c2The starting and stopping moments of the disturbance can be positioned, and the positioning precision can be comparable to that of a classical wavelet method.
To f1White noise with a signal-to-noise ratio of 20dB is added in (t), and if the disturbance is continuously detected in the spectrum division manner in FIG. 3, the decomposition result is shown in FIG. 4 (a). At this time, component c1、c2Are polluted by serious noise, and effective information is difficult to obtain from the noise. Therefore, for the disturbance detection problem in the noise environment, the invention corrects the boundary extension coefficient, and the adjusted spectrum division mode and the disturbance detection result are respectively shown in fig. 4(b) and (c). From this, the noisy signal is decomposed into a fundamental frequency component c1And a noise component c2Component c1Although the disturbance mutation information is lost, the amplitude characteristic is kept. C is obtained by HT and NHT respectively1The result of the instantaneous amplitude function shows that the amplitude of the NHT at the end point is more stable, and the flying wing phenomenon does not exist, so that the method is more suitable for extracting the disturbance information in the IEWT decomposition result.
As can be seen from FIG. 4(c), c1The instantaneous amplitude curve a (t) of (a) not only contains the disturbance amplitude information, but also can locate the disturbance moment. Considering that the rate of change of curve a (t) always reaches a maximum value near the midpoint of the rising or falling segment, the threshold τ is set to [ max (a) + min (a)]And/2, the positions of A (t) when the A (t) passes through the tau, namely the occurrence and termination time of the corresponding disturbance, are obtained from 0.0654s and 0.1649s, and the average error is only 0.3 ms. And processing the amplitude curve in the disturbance period by adopting a mean value fitting method after partial end points are removed, so that 0.4981 of disturbance amplitude can be obtained, and the detection error is less than 0.2%.
Example 2:
to further illustrate the real-time performance of the detection algorithm of the present invention, HHT, VMD, MIST and the method of the present invention are used to detect the pure disturbance signal f on the experimental computer equipped with i5-8500CPU and 16GRAM, respectively1(t) detection is carried out, and the time consumption comparison result of the algorithm is shown in the table 1.
Table 1 detection algorithm real-time comparison
As can be seen from Table 1, the VMD can analyze all 5 types of disturbances at the same time, but the algorithm takes the longest time due to the existence of multiple iterative decomposition processes; although MIST is simplified on the basis of original S transformation, the operation time is still long; HHT is relatively small in calculation amount, but cannot analyze complex disturbance; in contrast, the algorithm is based on wavelet transformation theory, iterative screening of disturbance signals is not needed, time consumption of the algorithm is about one tenth of HHT, and the algorithm has good real-time detection performance.
Example 3:
taking a single-phase short-circuit event occurring in a 35kV distribution line in a certain place as an example, the waveform of the short-circuit phase voltage during a fault period (after normalization, the same applies below) and the result of frequency spectrum division are shown in fig. 5, the sampling frequency of the wave recording device is 12.8kHz, and the signal duration is 0.24 s. IEWT time-frequency analysis is performed on the measured signal, and the disturbance decomposition result and its time-frequency spectrum are shown in fig. 6. As can be seen from fig. 6, IEWT effectively extracts the fundamental wave and the harmonics from the original signal by virtue of its good modal decomposition performance. Further observing the NHT time-frequency spectrogram drawn based on each disturbance component, the method can find that each modal amplitude is stable and single in frequency, and the amplitude and the frequency of the harmonic wave can be accurately obtained from the modal amplitude. In contrast, modal aliasing occurs in both the HHT and LMD methods because the amplitude-frequency parameters of each harmonic in the true harmonic perturbation are not constant, resulting in uneven distribution of the extreme points in the perturbation waveform.
In order to verify the anti-noise performance of the algorithm used for actually measured signal analysis, the method is adopted to analyze the disturbance under the conditions of no noise and 30dB signal-to-noise ratio respectively so as to verify the anti-noise performance of the algorithm used for actually measured signal analysis, and the result is compared with the result obtained by the IEC recommended harmonic measurement method based on windowed Fourier transform under the noise-free environment, and the result is shown in Table 2. The detection result of the method is very close to the IEC standard result, the average detection errors of the amplitude and the frequency are 0.68% and 0.07% respectively, the detection precision is good, and meanwhile, the method can accurately analyze the actual measurement disturbance even under the noise interference due to the strong noise tolerance.
TABLE 2 actual measurement of harmonic detection results
Claims (9)
1. A power quality disturbance signal detection method based on improved empirical wavelet transform is characterized by comprising the following steps:
s1: carrying out scale space representation on the frequency spectrum of the power quality disturbing signal, and determining the characteristic frequency point of the disturbing signal according to the dynamic measurement of the scale space representation result;
s2: extending the boundary on the basis of the original frequency band boundary, and constructing an experience wavelet on the basis of the extended frequency band boundary according to the acquired characteristic frequency points of the disturbance signal;
s3: decomposing the power quality disturbance signal into a plurality of amplitude modulation-frequency modulation components by using improved empirical wavelet transform;
s4: and acquiring a detection result of the disturbance signal according to the amplitude modulation-frequency modulation component.
2. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform as claimed in claim 1, wherein said step S1 specifically includes the following steps:
a1: setting an electric energy quality disturbing signal containing N characteristic frequency points as F (t), and performing fast Fourier transform on the signal to obtain a result F (omega);
a2: defining a discrete form of a scale-space representation of F (ω);
a3: and carrying out trend removing evaluation on the maximum value point in the scale space representation by adopting dynamic measurement, and determining the characteristic frequency point of the disturbance signal.
3. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform of claim 1, wherein the discrete form of the scale space representation of F (ω) in the step a2 is as follows:
in the formula: n is a scale factor and M is an approximation coefficient.
4. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform as claimed in claim 3, wherein said step A3 is specifically as follows: let xuIs a maximum point in L (ω, n) if there is a ratio x in L (ω, n)uHigher points, then xuDynamic measure D ofyn(xu) Equal to its dynamic measure to the smallest path among all the equal altitude point paths, i.e.:
Dyn(xu)＝{inf{Dyn[l(xu,xv)]}；halt(xu)＝halt(xv)}
in the formula: l (x)u，xv) Denotes xuTo the isocenter xvA path of (a); i.e. inf{. denotes infimum bound; h isalt(. cndot.) represents height; at this time, if Dyn(xu) Greater than a discrimination threshold TdCan determine xuIs a characteristic frequency point.
5. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform as claimed in claim 1, wherein the specific process of step S2 is as follows:
b1: capture all N characteristic frequency points { omeganAfter the Nn is equal to 1, selecting a frequency corresponding to a midpoint between two adjacent characteristic frequency points as a frequency band division boundary; at the same time, the first N boundaries ω are preservednAnd (5) extending the band boundary according to the following formula on the basis that N-1N is 0:
in the formula: lambda is an adjustable boundary extension coefficient; at this time, the signal spectrum will be divided into N +1 consecutive intervals { Λ +n}N+1n＝1；
B2: defining an empirical wavelet tightening framework
And constructing an empirical scale function in the frequency domain
And empirical wavelet function
Respectively performing inner products on f (t) and the empirical scale function and the empirical wavelet function to obtain approximate coefficients
t) and detail coefficients
6. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform as claimed in claim 5, wherein in said step B1, the power quality disturbance is divided into single frequency disturbance and multiple frequency disturbance according to the number of characteristic frequency points in the signal spectrum; for single-frequency power quality disturbance, setting lambda (pi/2-omega)N)/(ΩN-ωN-1) I.e. define the interval ΛN+1＝[π/2，π]Obtaining high-frequency distortion information contained in a disturbance signal frequency spectrum; when the disturbance signal is mixed with noise interference, considering that the disturbance component frequency spectrum is symmetrical about a peak point in a frequency domain, making λ ═ 1 can make the highest frequency component completely fall into an upper cut-off frequency and a lower cut-off frequency which are respectively ω while keeping the frequency domain characteristics of the disturbance signalN-1And ωNThereby being free from high frequency noise.
7. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform of claim 1, wherein the power quality disturbance signal in the step S3 is a voltage signal or a current signal.
8. The method for detecting the power quality disturbance signal based on the improved empirical wavelet transform as claimed in claim 1, wherein said step S4 specifically comprises: and (3) performing standard Hilbert transform on each disturbance component to obtain the instantaneous amplitude and instantaneous frequency of the disturbance component, so as to obtain a time-frequency spectrogram of the disturbance signal, and parameters of the disturbance amplitude, the disturbance frequency and the disturbance starting and stopping time.
9. The method for detecting power quality disturbance signal based on improved empirical wavelet transform as claimed in claim 8, wherein said method for calculating instantaneous amplitude and instantaneous frequency of disturbance component in step S4 is as follows:
c1: fitting the maximum value point of | x (t) | by adopting a cubic spline function to obtain an empirical envelope function a (t), and standardizing the single-component signal x (t) according to the following formula:
x1(t)＝x(t)/a(t)
c2: obtaining a normalized signal | x1(t) | empirical envelope function a1(t), iterating the steps for n times until an(t) is less than or equal to 1; to this end, the single component x (t) has been decomposed into the form of a product of amplitude modulation and frequency modulation, the resulting signal x being normalizedn(t) is the frequency modulated signal, and the amplitude modulated portion A (t) is defined as:
A(t)＝x(t)/xn(t)＝a(t)a1(t)…an(t)
c3: at this time, xn(t) already for an approximately pure frequency-modulated signal, its instantaneous frequency function f (t) is obtained directly by the hilbert transform:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110464981.5A CN113176460A (en) | 2021-04-28 | 2021-04-28 | Power quality disturbance signal detection method based on improved empirical wavelet transform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110464981.5A CN113176460A (en) | 2021-04-28 | 2021-04-28 | Power quality disturbance signal detection method based on improved empirical wavelet transform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113176460A true CN113176460A (en) | 2021-07-27 |
Family
ID=76926965
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110464981.5A Pending CN113176460A (en) | 2021-04-28 | 2021-04-28 | Power quality disturbance signal detection method based on improved empirical wavelet transform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113176460A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115048957A (en) * | 2022-06-08 | 2022-09-13 | 福州大学 | Composite power quality disturbance parameter identification method based on ISVMD-HT |
| CN116539333A (en) * | 2023-06-16 | 2023-08-04 | 中国汽车工程研究院股份有限公司 | Collision measurement signal preprocessing method and device based on vehicle impact sensor |
| CN116701822A (en) * | 2023-08-07 | 2023-09-05 | 云南电网有限责任公司 | Load electric energy metering method based on sparrow search algorithm and variation modal decomposition |
| CN116539333B (en) * | 2023-06-16 | 2024-07-16 | 中国汽车工程研究院股份有限公司 | Collision measurement signal preprocessing method and device based on vehicle impact sensor |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108732440A (en) * | 2018-04-17 | 2018-11-02 | 北京理工大学 | A kind of transient power quality detection method and system |
| CN108921082A (en) * | 2018-06-28 | 2018-11-30 | 哈尔滨工程大学 | Diesel combustion resonant frequency extracting method based on experience small echo and Hilbert transform |
| CN110967599A (en) * | 2019-12-11 | 2020-04-07 | 贵州电网有限责任公司 | Electric energy quality disturbance detection and positioning algorithm |
| CN111141520A (en) * | 2020-02-24 | 2020-05-12 | 江南大学 | Rolling bearing fault diagnosis method based on improved experience wavelet transform |
| CN111623982A (en) * | 2020-06-15 | 2020-09-04 | 大连理工大学 | Planetary gearbox early fault diagnosis method based on APEWT and IMOMEDA |
| CN112630527A (en) * | 2020-12-04 | 2021-04-09 | 云南民族大学 | Distortion signal electric quantity measuring method based on empirical wavelet transform |
-
2021
- 2021-04-28 CN CN202110464981.5A patent/CN113176460A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108732440A (en) * | 2018-04-17 | 2018-11-02 | 北京理工大学 | A kind of transient power quality detection method and system |
| CN108921082A (en) * | 2018-06-28 | 2018-11-30 | 哈尔滨工程大学 | Diesel combustion resonant frequency extracting method based on experience small echo and Hilbert transform |
| CN110967599A (en) * | 2019-12-11 | 2020-04-07 | 贵州电网有限责任公司 | Electric energy quality disturbance detection and positioning algorithm |
| CN111141520A (en) * | 2020-02-24 | 2020-05-12 | 江南大学 | Rolling bearing fault diagnosis method based on improved experience wavelet transform |
| CN111623982A (en) * | 2020-06-15 | 2020-09-04 | 大连理工大学 | Planetary gearbox early fault diagnosis method based on APEWT and IMOMEDA |
| CN112630527A (en) * | 2020-12-04 | 2021-04-09 | 云南民族大学 | Distortion signal electric quantity measuring method based on empirical wavelet transform |
Non-Patent Citations (1)
| Title |
|---|
| 吴建章 等: ""基于改进经验小波变换的电能质量扰动检测新方法"", 《电力自动化设备》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115048957A (en) * | 2022-06-08 | 2022-09-13 | 福州大学 | Composite power quality disturbance parameter identification method based on ISVMD-HT |
| CN116539333A (en) * | 2023-06-16 | 2023-08-04 | 中国汽车工程研究院股份有限公司 | Collision measurement signal preprocessing method and device based on vehicle impact sensor |
| CN116539333B (en) * | 2023-06-16 | 2024-07-16 | 中国汽车工程研究院股份有限公司 | Collision measurement signal preprocessing method and device based on vehicle impact sensor |
| CN116701822A (en) * | 2023-08-07 | 2023-09-05 | 云南电网有限责任公司 | Load electric energy metering method based on sparrow search algorithm and variation modal decomposition |
| CN116701822B (en) * | 2023-08-07 | 2023-11-28 | 云南电网有限责任公司 | Load electric energy metering method based on sparrow search algorithm and variation modal decomposition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112395992B (en) | Power harmonic signal denoising method based on improved wavelet threshold | |
| CN103245832B (en) | Based on harmonic wave time-frequency characteristic method for parameter estimation and the analyser of quick S-transformation | |
| CN113176460A (en) | Power quality disturbance signal detection method based on improved empirical wavelet transform | |
| CN108152674B (en) | Fault traveling wave filtering method based on feature point identification and linear interpolation | |
| CN107543962B (en) | Calculation method of dominant inter-harmonic frequency spectrum distribution | |
| Yao et al. | Fast S-transform for time-varying voltage flicker analysis | |
| CN111693775A (en) | Harmonic detection method, device and medium for power transmission network | |
| CN116865269B (en) | Wind turbine generator system high harmonic compensation method and system | |
| CN113325277A (en) | Partial discharge processing method | |
| Huda et al. | Power quality signals detection using S-transform | |
| CN113032716A (en) | Harmonic and inter-harmonic analysis method based on windowed interpolation and Prony algorithm | |
| Abidullah et al. | Real-time power quality signals monitoring system | |
| Chang et al. | A synchrosqueezing transform-based hybrid method for voltage fluctuations assessment | |
| Gupta et al. | Estimation of instantaneous frequencies using iterative empirical mode decomposition | |
| CN110095691B (en) | Method and device for extracting initial traveling wave head based on full-waveform main frequency component | |
| Hochbrückner et al. | Digital filtering methods for interferences on partial discharges under DC voltage | |
| Abdullah et al. | Bilinear time-frequency analysis techniques for power quality signals | |
| CN106872777B (en) | Harmonic and inter-harmonic separation analysis method | |
| CN114487589A (en) | Power grid broadband signal self-adaptive measurement method, device and system | |
| CN112034253B (en) | MOA online monitoring method | |
| CN112034232A (en) | Power supply system voltage sag detection method | |
| CN114184838A (en) | Power system harmonic detection method, system and medium based on SN mutual convolution window | |
| Ramos et al. | DSP based power quality analyzer using new signal processing algorithms for detection and classification of disturbances in a single-phase power system | |
| Xiao | A new wavelet coefficients correlation denoising method applied in fault detection | |
| Khan | Signal processing techniques used in power quality monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210727 |